Wizard's Toolkit has files prepared for common cron job operations. By default the cron jobs are located in a /cron folder which is public on the webserver. You can (and probably should) move that to a non-public location on the server. Since Nginx, Apache and other webservers have different methodologies for calling CRON jobs, those details can be decided by your DevOps person.

All cron job files are kept in the /cron/ folder. Of course, you can move these to a different folder as needed.

WTK has a default robots.txt which tells search engines to ignore the /cron/ folder and other folders which generally should not be crawled, for example the /exports/ folder. This robots.txt should be uploaded to your root directory of your web server. Modify it as needed for your specific needs.

To prevent calls from outside sources to your CRON jobs, a `pw` parameter is required. The password is specific to your server and you set what you want the password to be in the /wtk/wtkServerInfo.php file.

$gloAuthStatus = 'setYourUniqueCodeHere';  // guarantees uniqueness for login security level checks

Whatever value you set there, prepend 'wtk' before it when calling your cron jobs. For example, to trigger the monthly.php cron job you would call it this way:

https://your-domain.com/cron/monthly.php?pw=wtksetYourUniqueCodeHere

Different OS have different calling methods so check with your DevOps and verify the settings used work as expected and meet your security requirements. This is just one possible way of defining a cron job to call a specific web page:

/usr/bin/curl --silent --insecure --output - https://your-domain.com/cron/minute.php?pw=wtksetYourUniqueCodeHere > /dev/null 2>&1

These pages provide a solid starting point for any client. Each client's needs are going to be very specific, but using these as a starting point gives you a solid framework.

This page is called by all cron job pages. Here is where it checks the security and you can enhance the security as-required. If the calling page sets $gloDebug = 'Y' before the require('cronTop.php') it adds debug code:

if ($gloDebug == 'Y'):
    error_reporting(E_ALL | E_STRICT);
    ini_set('display_errors', 1);
    $pgSQL = 'INSERT INTO `wtkDebug` (`DevNote`) VALUES (:DevNote)';
    $pgSqlFilter = array('DevNote' => $gloMyPage . ' called');
    wtkSqlExec($pgSQL, $pgSqlFilter);
endif;

This page also starts a timer so during testing you can see how long the cron jobs take.

This page is called at the bottom of cron job pages. If there is $gloDebug information, it displays it. Then it shows how long the page took to process.

if ($gloDebug != ''):
    $pgHtm .= '<hr><h4>Debug</h4>' . $gloDebug . '<hr>' . "\n";
//  $pgHtm .= '<hr>' . $pgSQL . '<hr>' . "\n";
endif;
 
$pgTimePassed = (((hrtime(true) - $pgStartTime)/1e+6)/1000);
$pgHtm .= '<br><p>Time it took: ' . $pgTimePassed . ' seconds.</p>' . "\n";
 
wtkSearchReplace('col m4 offset-m4 s12','col m6 offset-m3 s12');
wtkMergePage($pgHtm, 'CRON Jobs', _WTK_RootPATH . '/htm/minibox.htm');

This cron job should be called on 1st of each month. It inserts a row into the `wtkIncomeByMonth` data table based on the `wtkRevenue` data for the prior month. This `wtkIncomeByMonth` data is used for moneyHistory.php analytic reporting.

After you modify it, this cron job would be called nightly and usually should be set to run some time between 1am and 5am, depending on your client's needs.

Currently the code has an example of how it could be used for processing monthly payroll on the 16th of the month. This way you can easily change it for your needs, whether that be triggering payroll on a certain day of the month or other nightly activities.

This file has the most functional code in it. All source code is available (as always) so you can easily modify and expand upon it for your needs.

• files more than 15 minutes old are deleted from /exports/ folder
• wtkNotifications processed and sent via Email or SMS
• migration of files from local storage to AWS S3 or Cloudflare R2 (code commented out)

If you are not using the `wtkNotifications` system, this will have no affect because will be a quick loop through a zero-row table.

If you are using AWS S3 or Cloudflare R2, with Wizard's Toolkit you can have the files uploaded to a local folder (or AWS EFS) initially since this is fastest. Then once per minute this minute.php cron job will loop through the local files (stored in `wtkFiles` data table), and migrate them to the external storage location. After it does that, it changes a flag in `wtkFiles` so thereafter when users access the file it pulls the file from S3 or R2. The settings for your S3 or R2 keys are all in wtk/wtkServerInfo.php.

These can easily be managed in the data table aptly named `wtkBackgroundActions`.